pdmphoto wrote:
You will see noticeable differences, in the extreme corners, with all these lenses stopped down to f8-11. Some don't have it no matter the aperture, the others have varying degrees (and types) of weakness, even when stopped down.
.
[
I agree. When I compare my CZ 35-70 to the Summicron 35/2 the corners and edges do resolve differently. Both are excellent lenses, but as an example the corner resolution roll-off on the Summicron starts closer to the extreme corner than the CZ. If I had to say which had more roll-off at the extreme corner I would say it is the Summicron. On the other hand, if I look at the extreme left and right hand edges of the frame (away from the corners) I think the Summicron has slightly better resolution. All of my comments are in the context of f/8 and near infinity focus - meaning landscape type set-up conditions.
B Landau wrote:
I agree. When I compare my CZ 35-70 to the Summicron 35/2 the corners and edges do resolve differently. Both are excellent lenses, but as an example the corner resolution roll-off on the Summicron starts closer to the extreme corner than the CZ. If I had to say which had more roll-off at the extreme corner I would say it is the Summicron. On the other hand, if I look at the extreme left and right hand edges of the frame (away from the corners) I think the Summicron has slightly better resolution. All of my comments are in the context of f/8 and near infinity focus - meaning landscape type set-up conditions....Show more →
Precisely. One reason among others why a test that only looks at 2 points is inadequate, and potentially misleading.
Precisely. One reason among others why a test that only looks at 2 points is inadequate, and potentially misleading."
Agreed, but then AG shows lots of passion, he is open to critics and he has improved his review site constantly.
He is also among the few which provides full size images of some of the tested lenses. I would prefer to have the possibility to download an interior or landscape photograph though. We probably agree that this much better than posting 3, 4 or 7 crops from one image.
pdmphoto wrote:
Nice work, but again the definition of "borders" seems to cloud the results. I've tried about half of the listed lenses and my corner results don't match these border results, so I'd be careful about interpretting the results from the "borders".
Valid question - take a look at the targets displayed in FAQ. MTF is measured off left top and right bottom slanted squares. These are 85% distance from the center (for self-assembled chart) and 80% distance from the center (for the iso chart).
pdmphoto wrote:
Besides the "borders" issue. I am at a loss to understand how these tests can be so different from my own real world tests. I'm sure it is a combination of factors, but I wonder if these tests are done at a distance considerably less than infinity? I've found some lenses do very well at distances close in, but progressively worse towards infinity. If that is the case than these tests are more close focus, than real world performance for the type of photography I do.
I think I have mentioned this in my other posts and it should definitely be in the FAQ. I record best MTF, meaning that I don't take one set at just a particular distance, but rather vary the distance to the chart as well to measure MTF at various distances. Thus one lens might be optimized for closeups while the other will be optimized for infinity - I will have the best MTF for both. I do not record at what distance the lens produced the best MTFs - just the best MTF out of all sequences. The process is relatively streightforward, it just requires bunch of measurements - start at a distance, go through focus brackets for all tested apertures, move the camera/lens by 1m, repeat, move by 1m, repeat etc...
This is all described in the FAQ section - specifically put it there since I was repeatedly getting this question.
B Landau wrote:
With all the positive comments I have beeen reading about the ZF 35/2 I am surprised it did not fair somewhat better in your lab tests - particularly with regard to edge resolution.
The summicron 35/2 is no doubt a very good lens, but on my copy I do see noticeable resolution loss in the extreme most corner regions (when shooting at f/8). Having said that the edge resolution of non-corner regions is extremely good.
well, i will post the real life test results in a day or two - i have all pictures taken, it's just a matter of massaging them to fit the web page and adding colorful comments that sometimes create a lot of controversy ;-) seriously, i'll have full sized images uploaded as well, not just crops - it would be quite interesting to see whether the real life tests back MTF measurements...
Arianne Dubois wrote:
Contax 35-70 (in our tests the best 35mm lens)
Canon EF 24-105 f/4 (astonishing at 35mm and better than 35 f/1.4)
Contax 35 PC (second best Contax 35mm)
These are great lenses, and I happened to have two of them (Canon 24-105l and Contax 35-70). At some point I thought throwing in a bunch of zooms in as well, but then decided to keep it clean and limit only to fixed focals. I will be scheduling a test for zooms at a later time...
Andi Dietrich wrote:
He is also among the few which provides full size images of some of the tested lenses. I would prefer to have the possibility to download an interior or landscape photograph though. We probably agree that this much better than posting 3, 4 or 7 crops from one image.
I promise to start posting more full samples taken during the tests - I will probably have to migrate to a larger service and bandwidth but that's ok since I was planning to do that anyway. I will also stop posting artistic type photos (i.e. HDR etc...)
If somebody wants to add to the sample galleries, you can do it now - just register and go into upload sample image section... Your samples will show up in the lens section which you select...
I think most of us realize that no testing methodology is going to reveal all the strengths and weaknesses of a given lens, but I do think that the data that eosslr provides is useful. I for one appreciate the efforts he has made to provide some level of quantification to balance the many subjective comments.
eosslr wrote:
These are great lenses, and I happened to have two of them (Canon 24-105l and Contax 35-70). At some point I thought throwing in a bunch of zooms in as well, but then decided to keep it clean and limit only to fixed focals. I will be scheduling a test for zooms at a later time...
That would be great. Thanks for all of the effort on your behalf.
